Short-Term Trading by Retail Investors

Tue, 07 Jul 2026 21:55:00 +0000

The empirical evidence from Brazil and the European Union: the facts for a clear-eyed decision about the risk of loss

Scope: day trading, contracts for difference (CFDs) and binary options.

1. Introduction

Short-term trading, called day trading in the financial literature when positions are opened and closed within the same day, is marketed by sellers of trading courses and by brokerage platforms as an accessible way for retail investors to earn an income. Two independent sources, one academic and one regulatory, make it possible to test that claim against observed data. The first is a study by three Brazilian researchers covering every individual who took up day trading in Brazil’s equity futures market between 2013 and 2015. The second is the body of data gathered by the national supervisory authorities of several European Union Member States and compiled by the European Securities and Markets Authority (ESMA), the evidence behind the first pan-European intervention on financial products sold to retail investors. This note describes both sources, sets out the quantitative results, explains the mechanisms that drive the losses and the regulatory consequences, including the measures adopted in Portugal by the securities regulator, the CMVM, and gives the reader what is needed for a clear-eyed, fact-based decision.

97%	74% to 89%	€1,600 to €29,000
of the retail traders who persisted at day trading for more than 300 days in Brazil’s equity futures market (2013 to 2015) lost money	of retail CFD accounts lost money, according to the supervisory authorities of several European Union Member States	the range of average losses per retail CFD client established by those authorities and compiled by ESMA

2. The Brazilian study: methodology and scope

The study, titled Day Trading for a Living?, is the work of Fernando Chague and Rodrigo De-Losso, of the Department of Economics at the University of São Paulo’s School of Economics, Business and Accounting (FEA-USP), and Bruno Giovannetti, of the São Paulo School of Economics at the Getulio Vargas Foundation (FGV EESP). A first version circulated as working paper 2019-47 of the FEA-USP Department of Economics, dated 19 August 2019. The revised version, dated 11 June 2020, is available on the Social Science Research Network as SSRN 3423101 and also appeared as FGV EESP discussion paper (Texto para Discussão) 525.

The authors tracked every individual who began day trading Brazilian equity futures between 2013 and 2015, in what was then the world’s third-largest market by volume of equity futures contracts. The dataset covers the entire population of new traders over that period, not a sample, which removes the self-selection bias that besets voluntary surveys. The core of the analysis is the group who stayed in the market for at least 300 trading days, precisely the people who, on the course sellers’ promise, should have been reaping the benefit of accumulated experience.

3. The Brazilian study: results

Of those who persisted past 300 days, 97% lost money. In the June 2020 version of the study, just 1.1% made more than the Brazilian minimum wage, and only 0.5% made more than a bank teller’s starting salary, in every case while running high risk. The best performer in the entire population averaged a gain of 310 US dollars a day with a daily standard deviation of 2,560 dollars, results more than eight times as volatile as the gain itself. The August 2019 version had drawn the comparison threshold slightly differently, with 0.4% of participants earning more than a bank teller’s daily income, estimated at 54 dollars a day.

The study adds a finding that bears directly on the idea of learning by doing. The authors found no evidence that performance improved with time in the market. Persistence did not turn losing traders into profitable ones, which contradicts the marketing claim that accumulated experience leads to proficiency. In the 2020 version, the authors set this result in the context of the structural competition that retail traders face from high-frequency trading systems, and concluded that making a living from day trading is virtually impossible.

4. The European evidence: what the national authorities found

In Europe, the quantitative evidence was produced by the national authorities responsible for supervising financial markets in several Member States and compiled by ESMA in the course of its product intervention on contracts for difference (CFDs) and binary options. Their analyses of CFD trading across European Union jurisdictions showed that between 74% and 89% of retail accounts lose money on these products, with average losses per client of €1,600 to €29,000. In Spain specifically, the Comisión Nacional del Mercado de Valores found that 82% of the clients who traded CFDs between 1 January 2015 and 30 September 2016 took losses, €142 million in all.

These figures concern contracts for difference, leveraged derivatives in which the investor never acquires the underlying asset and simply contracts with the broker to exchange the difference between a position’s opening and closing prices. Leverage magnifies gains and losses alike, and frequent trading multiplies the transaction costs the investor bears. The convergence between the European range, 74% to 89% of accounts in loss, and the 97% of losers in the Brazilian study strengthens the pattern, since the two sources rest on different populations, instruments, markets and methods.

5. ESMA’s intervention: legal basis, measures and dates

On the strength of these data, ESMA’s Board of Supervisors approved the first product intervention measures on 23 March 2018, under Article 40 of Regulation (EU) No 600/2014 (MiFIR), a power that allows it to restrict temporarily, for renewable three-month periods, the marketing of financial instruments that raise significant investor protection concerns. The formal decision on CFDs was adopted on 22 May 2018, and the two decisions, one on binary options and one on CFDs, were published in the Official Journal of the European Union on 1 June 2018 under references CELEX 32018X0601(01) and 32018X0601(02).

The ban on marketing, distributing or selling binary options to retail investors took effect on 2 July 2018. The restrictions on CFDs took effect on 1 August 2018 and have five components. The first sets leverage caps on the opening of positions, graduated by the volatility of the underlying asset, as in the table below. The second requires positions to be closed out automatically once the account’s value falls below 50% of the initial margin required. The third guarantees negative balance protection per account, capping the investor’s loss at the money deposited. The fourth bans incentives to trade CFDs. The fifth requires every broker to publish, in standardised form and updated quarterly, the percentage of its retail accounts that lost money over the previous twelve months.

Maximum leverage for positions opened by retail investors (ESMA Decision of 22 May 2018):

Cap	Underlying
30:1	major currency pairs
20:1	non-major currency pairs, gold and major equity indices
10:1	other commodities and non-major indices
5:1	individual shares and other underlyings
2:1	cryptocurrencies

6. Making the restrictions permanent in Portugal: the CMVM’s response

Because MiFIR made ESMA’s measures temporary by nature, entrenching them required national regulation. The CMVM put out to public consultation, with a deadline of 27 February 2019, a draft regulation to make the European restrictions permanent in Portugal, banning the marketing, distribution and sale of binary options to non-professional investors and restricting the marketing of CFDs to that same public. The CMVM grounded the decision in the ESMA data described above and warned that not adopting the measures in Portugal would create room for regulatory arbitrage, with the supply of these products predictably migrating to the Portuguese market from jurisdictions where the restrictions were in force.

To this is added a tax regime with a direct effect on the net result of investors resident in Portugal. Gains on CFDs are taxed as capital gains on securities at a flat 28% and must be declared under category G of the personal income tax return (IRS), which further reduces the income of anyone who does manage to turn a profit.

7. Why the losses happen: the mechanisms documented in the sources

The two sources do more than quantify the losses. Both identify the mechanisms that produce them, and where the two converge, five documented causes can be isolated.

The first cause is the effect of leverage on the probability of losing the entire margin. ESMA’s technical analysis of 1 June 2018, published under reference ESMA50-162-215, quantifies the mechanism with a worked example. On a gold position leveraged at 100:1, the initial margin required for a contract with a notional value of €10,000 is €100. A 3% rise in the price of gold produces a 300% return on the initial margin, but a fall of just 1% wipes that margin out. The asymmetry does not lie in the odds of the price moving one way or the other, which are the same in both directions. It lies in the fact that losing the whole margin closes the position and forecloses any recovery, while an equivalent gain closes nothing. The same analysis concludes that high leverage raises the probability that a position will be closed out automatically within a short time, before any favourable turn in the price.

The second cause is transaction costs multiplied by trading frequency. Every trade bears the spread between the buying and the selling price, the broker’s commissions and, on leveraged instruments, the cost of funding the position. These charges are fixed per trade and indifferent to the outcome, so their total grows in direct proportion to the number of transactions. ESMA’s analysis of 1 June 2018 notes that retail investors typically hold CFDs only briefly, which means high capital turnover and, with it, fast-accumulating costs. Before any profit, a short-term trader has to generate a gross return large enough to cover charges that a long-term investor pays once.

The third cause is the structural conflict of interest in contracts for difference. ESMA’s decision of 22 May 2018 documents that, in the prevailing business model, the CFD broker is the direct counterparty to the client’s position, so one party’s profit is the other’s loss. The same decision identifies two further risks that flow from that position. The first is the asymmetric exploitation of slippage, in which the broker passes on to the client the loss arising from the gap between the quoted price and the executed price, and keeps the equivalent gain when the gap runs in its favour. The second is the deliberate delay between quote and execution, which widens the asymmetry. On binary options, ESMA went beyond the conflict of interest and concluded that the product’s expected return is negative by design: the average loss is built into the instrument before the investor makes a single decision, and that finding justified an outright ban rather than a mere restriction.

The fourth cause is the structural competitive disadvantage against professional operators. In the version of 11 June 2020, Chague, De-Losso and Giovannetti set their results in the context of the competition between retail investors and high-frequency trading systems. These systems run on servers co-located with the exchanges’ matching engines, operate at latencies measured in microseconds, and systematically capture price opportunities before any retail order reaches the market. A retail trader working in minutes competes in the same order book, on slower information and slower execution, against counterparties whose technological edge is permanent.

The fifth cause is the absence of learning, which keeps the early losses from ever being recovered. The Brazilian study tested directly whether performance improves with experience and found no evidence that persistence turned losing traders into profitable ones. That finding converts the four previous causes from passing costs into permanent ones. If learning existed, the losses of the first months would be an investment in future skill. In its absence, every additional day of activity repeats the same structure of disadvantage, and the cumulative probability of loss grows with time in the market instead of shrinking.

These five causes operate at once and reinforce one another. Leverage speeds up the loss of the margin, transaction costs eat into capital with every trade, the conflict of interest and the technology gap degrade the quality of execution, and the absence of learning ensures that experience neutralises none of it. The 74% to 89% of losing accounts in the European Union and the 97% of persistent losers in Brazil are the aggregate result of that combination, not of individually bad decisions taken in isolation.

8. In sum: the facts for the reader’s decision

The two sources arrive by independent routes at the same quantified conclusion. In Brazil’s equity futures market, 97% of the retail traders who persisted at day trading for more than 300 days between 2013 and 2015 lost money, with no evidence of learning, according to the study by Chague, De-Losso and Giovannetti. In the European Union, the national supervisors’ analyses compiled by ESMA put the share of retail CFD accounts in loss at between 74% and 89%, with average losses of €1,600 to €29,000 per client. The European regulatory response of 2018, and its Portuguese consolidation in 2019, rested on this empirical basis and on nothing else.

The facts that matter for the decision are therefore these. The documented probability of loss runs from 74% to 97%, depending on the market and the instrument. The mechanisms that produce the losses are structural, operate simultaneously and, on the available evidence, are not neutralised by experience. The European and Portuguese regulators judged the risk serious enough to restrict the marketing of these products and to ban binary options outright. The documented pattern concerns short-term trading in leveraged instruments and does not carry over, without separate analysis, to long-term equity investing, whose costs, risks and time horizon are of a substantially different order. This note is not individual financial advice. Anyone weighing this activity decides in full knowledge that the expected outcome, measured across entire populations of real traders, is negative for the overwhelming majority.

9. Sources

Chague, F., De-Losso, R., and Giovannetti, B., Day Trading for a Living?, working paper 2019-47, Department of Economics, FEA-USP, 19 August 2019; revised version of 11 June 2020, SSRN 3423101 (https://ssrn.com/abstract=3423101); also published as Texto para Discussão 525, FGV EESP, 2020.

ESMA, press release “ESMA agrees to prohibit binary options and restrict CFDs to protect retail investors”, reference ESMA71-98-128, 27 March 2018, recording the Board of Supervisors’ approval of 23 March 2018.

ESMA, press release “ESMA adopts final product intervention measures on CFDs and binary options”, 1 June 2018, giving the application dates of 2 July 2018 for binary options and 1 August 2018 for CFDs.

ESMA decisions published in the Official Journal of the European Union on 1 June 2018, references CELEX 32018X0601(01) and 32018X0601(02), adopted under Article 40 of Regulation (EU) No 600/2014.

ESMA, “Product Intervention Analysis: Measure on Contracts for Differences”, reference ESMA50-162-215, 1 June 2018, referring to the decision of 22 May 2018 on CFDs.

CNMV International Bulletin, “ESMA agrees on product intervention measures in relation to CFDs and binary options offered to retail investors”, June 2018, with the national authorities’ data (74% to 89% of accounts in loss, average losses of €1,600 to €29,000) and the Spanish finding by the CNMV (82% of clients in loss and €142 million between 1 January 2015 and 30 September 2016).

Jornal de Negócios, “CMVM impõe limites à comercialização de CFD a particulares”, 1 February 2019, and Jornal Económico, “CMVM limita a venda do derivado CFD a investidores não profissionais”, reporting the CMVM public consultation that closed on 27 February 2019.

cfd · npcrf.org